Open problems in human trait genetics

Abstract

Genetic studies of human traits have revolutionized our understanding of the variation between individuals, and yet, the genetics of most traits is still poorly understood. In this review, we highlight the major open problems that need to be solved, and by discussing these challenges provide a primer to the field. We cover general issues such as population structure, epistasis and gene-environment interactions, data-related issues such as ancestry diversity and rare genetic variants, and specific challenges related to heritability estimates, genetic association studies, and polygenic risk scores. We emphasize the interconnectedness of these problems and suggest promising avenues to address them.

Keywords: Causal variants; Complex human traits; Diversity; Epistatis; GWAS; Gene-environment interactions; Genome-wide association studies; GxE; GxG; Heritability; Human phenotypes; Linkage disequilibrium; Missing heritability; Non-additive genetic effects; PRS; Polygenic risk scores; Population structure; Rare variants; Recessive effects; Statistical genetics.

Fig. 1

Population structure confounds human genetic studies. A The population that an individual is born into influences their genetics and their environment, which are the two components affecting traits. As a result, genetic associations with human traits are confounded by population structure. B Even when considering a specific human group and controlling for the major axes of genetic variation in a cohort, the allele frequency of some variants can still vary across populations and exhibit clear geographic patterns, a problem known as “residual population structure”

Fig. 2

Identifying causal variants in the presence of linkage disequilibrium. A A single causal variant is in linkage disequilibrium with other nearby variants. As a result, variants that are correlated with the causal variant also obtain significant p-values even though they are not causal. B Combining GWAS summary statistics from three different ancestry groups, each exhibiting a different linkage disequilibrium pattern, to fine-map the results. By assuming that only one of the variants is causal, it can be recovered with high confidence

Fig. 3

Estimating heritability. Common methods for estimating the heritability of human traits. A In twin studies, heritability is estimated by the degree to which monozygotic (identical) twins are more phenotypically similar to each other than dizygotic (non-identical) twins. B In GREML, heritability is estimated by comparing genetic and phenotypic similarities across pairs of unrelated individuals. C In family-based methods, given a pair of individuals and their parents, the degree to which they are more genetically similar than would be expected from their parents can be compared to their phenotypic similarity to estimate the heritability of the trait